At present, a mainstream framework of a microwave communications device is a separated framework, that is, an analog part and a digital part of the device are separately located in two modules.
A digital module is configured to convert a signal received by the microwave communications device into a microwave digital signal, and output the microwave digital signal to an analog module after performing processing such as compensation on the signal.
The analog module is configured to convert, into a microwave analog signal, the microwave digital signal output by the analog module, and transmit the microwave analog signal after performing processing such as power amplification on the signal.
In order to reduce non-linear distortion of a power amplifier in the analog module, a DPD technology is generally used in the digital module to perform pre-calibration. That is, when a working state of the analog module changes, a DPD calibration coefficient in the digital module also needs to adaptively change at the same time in order to ensure a DPD calibration effect.
In the other approaches, a method for determining a DPD calibration coefficient is as follows. A DPD calibration coefficient corresponding to an extreme working state of an analog module is obtained in advance using an equipment and instrument environment shown in FIG. 1, and a DPD calibration coefficient corresponding to another working state of the analog module is calculated in a polynomial approximation manner. Details are as follows:
A function is preset:y=f(x)=k1x+k2x2+k3x3+k4x4+k5x5+ . . . ,where y is a DPD calibration coefficient corresponding to a working state x of the analog module, and k1, k2, k3, k4, k5, and the like are coefficients of the function.
The coefficients k1, k2, k3, k4, k5, and the like are determined, using a function iteration algorithm, based on the DPD calibration coefficient corresponding to the extreme working state of the analog module and that is obtained in advance, that is, function y=f(x) is determined such that a DPD calibration coefficient corresponding to each working state of the analog module can be calculated.
When an output power represents a working state of the analog module, if an output power range of the analog module is [PMin, PMax], the output power PMax of the analog module indicates an extreme working state of the analog module. Therefore, a corresponding DPD calibration coefficient when an output power of the analog module is PMax may be obtained in advance, and a corresponding DPD calibration coefficient when an output power of the analog module is another value, that is, a DPD calibration coefficient corresponding to another working state of the analog module, is calculated in the foregoing polynomial approximation manner. However, the DPD calibration coefficient obtained in this manner has a relatively large error, and as a result, a calibration effect cannot be ensured.